The present invention generally relates to a data processing apparatus having a list creating or generating function. More particularly, the invention is concerned with a list sort control method which allows list sort processing to be performed blockwise on a plural-row basis.
Of data processors such as word processors, personal computers and the like known heretofore, there exist those which are imparted with a function of creating or generating a list which contains data arrayed or arranged in the form of rows and columns, which data will hereinafter be referred to as list data. Many of the data processing apparatuses having the list creating function is imparted with a list data sorting function for determining relative magnitudes of values of the list data in a given row or column (hereinafter referred to as row data or column data) and rearranging the row data (or column data) in such a manner that the row data (or column data) are arrayed in an ascending order or sequence (i.e. in the order of small to large data values) or in a descending order (i.e. in the order or sequence of large to small data values). The column providing a basis for determining the magnitudes for the sequencing is referred to as the sort key which can be designated by the operator. In general, a column of numeric data is ordinarily designated as the sort key. Besides, a column of literal characters such as alphabet, kana (Japanese syllabary alphabet) or the like can also be designated as the key column. In the latter case, magnitudes or sequence of the literal characters is determined by making use of appropriate character codes.
Further, it is also practiced to designate simultaneously a plurality of columns as keys, wherein the sort is performed by assigning priority levels to the key columns. These keys are performed by assigning priority levels to the key columns. These keys are termed a first key, a second key, a third key and so forth. In the case, the sorting is so performed that the data or items of the first key are rearranged, wherein for the row data determined to be of a same level in the rearrangement performed for the first key, the sequence relation is determined by using the second key, while for the row data determined to be of a same level in the rearrangement performed with reference to the second key, the sequence relation is then determined by using the third key and so forth. In this manner, the sorting is performed by changing sequentially the keys from the high to low priority level.
As the prior art data processing techniques for the list data arrayed in rows and columns, there can be mentioned those disclosed in JP-A-60-220451, JP-A-62-254264 and JP-A-63-36476.
The prior art data processing apparatuses suffer from the problems in respect regard to the list data sorting function, which will be discussed below.
In the list data sorting known heretofore, a unit to be moved or manipulated for the rearrangement consists unexceptionally of one row (or one column), and determination of magnitudes or levels of the key data is performed on a row basis (i.e. row by row). There exists however the need for preparing or processing a list containing a meaningful set of data which are distributed among a plurality of rows. By way of example, let's suppose that list data shown in FIG. 18 of the accompanying drawings are to be sorted in accordance with the prior art sorting method. As can be seen in FIG. 18, there are assigned a plurality of rows to one bank, wherein each of the column labeled "BANK NAME" and the column labeled "BANK CODE" contains only one datum over a plurality of rows. To say in another way, there arise blank rows containing no data for the columns "BANK NAME" and "BANK CODE", respectively.
Accordingly, when the decision of the key data for magnitude or sequence thereof is performed on a row basis for the list data as shown in FIG. 18 to thereby rearrange the data row by row according to the prior art list data sorting method, the blank areas are regarded as the data. As a result the data for the different banks will mixedly be rearranged, giving rise to a problem. This problem can certainly be solved by placing the same data in the blank areas. (In the case of the list shown in FIG. 18, the blank rows of the columns "BANK NAME" and "BANK CODE" are all filled with the same bank name and the same bank code, respectively.) In that case, however, a new problem will arise in that the ease of seeing and understanding the list data is seriously disturbed or impaired.